Matt Flynn of Travelers gave a great half-day workshop Sunday on JMP, SAS and R. Attendees at the workshop were a mix of R, JMP and SAS users looking to augment their toolkits. Matt’s talk the following day on “Insurance and R” was very well-attended. One of my favorite takeaways was that using scripting as the glue that holds things together lets you take the best of what you like from different tools. You can then run longer processes in a more automated way and create more analytic bandwidth.
Koren Ichihara of Sears Holdings Corp. gave a very nice, concise presentation on HR analytics and squeezing insights from your people data using JMP: hire better people, retain top talent, inspire profitable behaviors. My favorite quote from her talk: “Visuals complement every analysis.” Dean Abbott also gave a talk on HR analytics, and the International Institute of Analytics member meeting that follows PAW features more on HR analytics.
John Elder’s“Data Mining Lessons Learned” featured real-world case studies, and even though I’ve heard some of the case studies before, I always learn something new in the way John relays these real-world examples.
Bob Grossman’s talk on analytic maturity was well attended and interesting. His first question was “How to speed up analytic maturity?” This reflects the greatest challenge – that of changing organizational culture. Thankfully, there are many who are happy to share their experiences at conferences like PAW, Discovery Summit, Analytics 2012 and in the growing number of analytical LinkedIn groups.
The JMP booth was situated across from the Central Intelligence Agency, which was at PAW to recruit data scientists. We discussed the challenges of attracting and retaining good talent in this area, including challenges with the BYOT (bring your own technology) Gen-Y group.
The term du jour, “big data,” of course came up, but James Taylor noted that most companies aren’t really taking advantage of the customer and process data they have, which aren’t always “big.”
Bruno Aziza, a former Microsoft business intelligence strategy director, noted that high-quality data visualization adds value to data analysis. “An image is always better than a tabular report,” as long as the graphics communicate analysis results clearly, Aziza said.
Max Kuhn’s talk, “Right Medicine, Right Patient” was excellent. I loved that he plugged Super Crunchersby Ian Ayres, who will be a keynote speaker at Discovery Summit 2012. And Max nicely summarized some common challenges: nontechnical experts’ acceptance of predictive models, disruptive technologies, visualization and presentation of results, and accuracy versus interpretability.
On the lighter side, JMP offered PAW attendees the opportunity to make their own “data movies” and take home flip books created from them. People loved them!